[concurrency-interest] Extended access methods for Atomics (and AQS)

Doug Lea dl at cs.oswego.edu
Wed Apr 21 07:34:44 EDT 2010


On 04/20/10 13:06, Joe Bowbeer wrote:
> I think it would be helpful to hear directly from more of the folks who
> use the Unsafe stuff currently.
>

You might not hear from too many of them on this list.
If you peruse google code search looking for usages,
you find (only) a dozen or so of them, including for example
Cliff Click's alternative forms of concurrent hash tables.
(Cliff's not on this list.)
Plus, there are those (as represented by the posting from
Benedict last week) who are frustrated not being able to use
efficient atomics or ordering control as they now stand.

Extrapolating, this might mean that there are
maybe a hundred or so total targeted usages, which is what
I meant by "small community". But this small community
includes many of those trying to create new efficient
concurrent components.

-Doug





More information about the Concurrency-interest mailing list